<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Saloni Tehariya</title>
    <description>The latest articles on DEV Community by Saloni Tehariya (@saloni_tehariya_350b45c25).</description>
    <link>https://dev.to/saloni_tehariya_350b45c25</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4243954%2F2be8f0e0-e584-4c09-96b5-c5028f062ea8.png</url>
      <title>DEV Community: Saloni Tehariya</title>
      <link>https://dev.to/saloni_tehariya_350b45c25</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/saloni_tehariya_350b45c25"/>
    <language>en</language>
    <item>
      <title>5 things interviewers listen for in a system design answer (that most candidates skip)</title>
      <dc:creator>Saloni Tehariya</dc:creator>
      <pubDate>Sat, 14 Nov 2026 06:42:15 +0000</pubDate>
      <link>https://dev.to/saloni_tehariya_350b45c25/5-things-interviewers-listen-for-in-a-system-design-answer-that-most-candidates-skip-1ei5</link>
      <guid>https://dev.to/saloni_tehariya_350b45c25/5-things-interviewers-listen-for-in-a-system-design-answer-that-most-candidates-skip-1ei5</guid>
      <description>&lt;p&gt;Most system design rejections aren't about picking the wrong database. They're about how the answer is built. After reviewing a lot of practice answers, the same five gaps keep coming up.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Jumping to components before constraints
&lt;/h2&gt;

&lt;p&gt;"I'd use Kafka" in minute one tells the interviewer you're pattern-matching. Spend the first 3–5 minutes on users, read/write ratio, latency targets and scale. Say the numbers out loud.&lt;/p&gt;

&lt;h2&gt;
  
  
  2. Listing tech without trade-offs
&lt;/h2&gt;

&lt;p&gt;Redis, sharding, a queue: fine, but why? For each choice, name one thing you gain and one thing you give up. For example, sharding by user ID spreads load but makes cross-user queries expensive.&lt;/p&gt;

&lt;h2&gt;
  
  
  3. No failure story
&lt;/h2&gt;

&lt;p&gt;Ask yourself what happens when this node dies, this queue backs up, or this cache goes cold. One sentence per component is enough to signal seniority.&lt;/p&gt;

&lt;h2&gt;
  
  
  4. Hedging language
&lt;/h2&gt;

&lt;p&gt;"I think", "probably" and "it depends" with nothing after them read as low ownership. Replace them with: "I'd choose X because Y; if Z changes, I'd switch to W."&lt;/p&gt;

&lt;h2&gt;
  
  
  5. No summary
&lt;/h2&gt;

&lt;p&gt;End by walking the request path once, end to end, and naming the biggest risk. It shows structure and gives the interviewer an easy place to probe.&lt;/p&gt;

&lt;h2&gt;
  
  
  How to practice this
&lt;/h2&gt;

&lt;p&gt;Record yourself answering one prompt ("design a URL shortener") out loud for 30 minutes. Replay it and tick off each of the five points. Do the same prompt again a few days later and compare.&lt;/p&gt;

&lt;p&gt;If you want structured feedback instead of self-review, I work on &lt;a href="https://gignix.com/?utm_source=devto&amp;amp;utm_medium=referral&amp;amp;utm_campaign=backlinks" rel="noopener noreferrer"&gt;Gignix&lt;/a&gt;, which runs AI mock interviews for coding, system design and behavioral rounds and gives a report on structure, depth and delivery. You can try one without signing up. But the recording method alone will get you most of the way.&lt;/p&gt;

&lt;p&gt;What's the gap you keep catching in your own answers?&lt;/p&gt;

</description>
      <category>career</category>
      <category>interview</category>
      <category>systemdesign</category>
      <category>beginners</category>
    </item>
  </channel>
</rss>
